Giter VIP home page Giter VIP logo

tiq's Introduction

Temporal Implicit Questions (TIQ)

Description

This code is for the TIQ benchmark proposed in our WWW'24 full paper "Faithful Temporal Question Answering over Heterogeneous Sources".

Please visit the following repo LINK to access the code for the FAITH method.

In this paper, we construct a new benchmark with focus on implicit temporal questions.

overview

This construction process operates as follows:

  • (i) sample a set of topic entities to start with;
  • (ii) retrieve temporal information snippets for each such topic entity from Wikipedia text, Wikipedia infoboxes, and Wikidata;
  • (iii) concatenate information snippets using a suitable temporal signal and construct an interrogative sentence, a pseudo-question;
  • (iv) rephrase the pseudo-question into a natural question using a generative model.

For more details see our paper: Faithful Temporal Question Answering over Heterogeneous Sources and visit our project website: https://qa.mpi-inf.mpg.de/tiq.

If you use this code, please cite:

@article{jia2024faithful,
  title={Faithful Temporal Question Answering over Heterogeneous Sources},
  author={Jia, Zhen and Christmann, Philipp and Weikum, Gerhard},
  journal={arXiv preprint arXiv:2402.15400},
  year={2024}
}

Environment setup

We recommend the installation via conda, and provide the corresponding environment file in environment.yml:

  git clone https://github.com/zhenjia2017/TIQ.git
  cd TIQ/
  conda env create --file environment.yml
  conda activate tiq
  pip install -e .

Alternatively, you can also install the requirements via pip, using the requirements.txt file.

Dependencies

TIQ makes use of CLOCQ for retrieving facts from WIKIDATA. CLOCQ can be conveniently integrated via the publicly available API, using the client from the repo.

Data

You need the following data. You can download from here:

  • wikipedia_wikidata_mappings.pickle
  • wikipedia_mappings.pickle
  • wikidata_mappings.pickle
  • types.pickle
  • labels.pickle
  • augmented_wikidata_mappings.pickle

Benchmark Construction Pipeline

To construct the benchmark requires following major steps.

1. Retrieve year pages

  bash scripts/pipeline.sh --year-page-retrieve <PATH_TO_CONFIG>

2. Construct pseudo-questions

  bash scripts/pipeline.sh --pseudoquestion-generate <PATH_TO_CONFIG>

3. Rephrase pseudo-questions

  bash scripts/pipeline.sh --question-rephrase <PATH_TO_CONFIG>

tiq's People

Contributors

zhenjia2017 avatar

Stargazers

Rishiraj Saha Roy avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.